A breakthrough astrophysics code, named Octo-Tiger, simulates the evolution of self-gravitating and rotating techniques of arbitrary geometry working with adaptive mesh refinement and a new process to parallelize the code to obtain excellent speeds.

This new code to model stellar collisions is additional expeditious than the founded code employed for numerical simulations. The analysis came from a unique collaboration involving experimental computer system experts and astrophysicists in the Louisiana Condition University Office of Physics & Astronomy, the LSU Center for Computation & Technology, Indiana University Kokomo and Macquarie University, Australia, culminating in above of a 12 months of benchmark screening and scientific simulations, supported by multiple NSF grants, which includes just one specially created to break the barrier involving computer system science and astrophysics.

“Thanks to a sizeable hard work throughout this collaboration, we now have a reliable computational framework to simulate stellar mergers,” stated Patrick Motl, professor of physics at Indiana University Kokomo. “By considerably decreasing the computational time to total a simulation, we can start off to inquire new issues that could not be dealt with when a one-merger simulation was important and really time consuming. We can investigate additional parameter room, take a look at a simulation at really superior spatial resolution or for for a longer time instances just after a merger, and we can increase the simulations to involve additional total actual physical versions by incorporating radiative transfer, for illustration.”

A short while ago printed in Every month Notices of the Royal Astronomical Society, “Octo-Tiger: A New, 3D Hydrodynamic Code for Stellar Mergers That Employs HPX Parallelisation,” investigates the code overall performance and precision as a result of benchmark screening. The authors, Dominic C. Marcello, postdoctoral researcher Sagiv Shiber, postdoctoral researcher Juhan Frank, professor Geoffrey C. Clayton, professor Patrick Diehl, analysis scientist and Hartmut Kaiser, analysis scientist, all at Louisiana Condition University — alongside one another with collaborators Orsola De Marco, professor at Macquarie University and Patrick M. Motl, professor at Indiana University Kokomo — as opposed their results to analytic remedies, when identified and other grid-centered codes, this kind of as the preferred FLASH. In addition, they computed the interaction involving two white dwarfs from the early mass transfer as a result of to the merger and as opposed the results with previous simulations of equivalent techniques.

“A check on Australia’s fastest supercomputer, Gadi (#25 in the World’s Leading five hundred record), showed that Octo-Tiger, operating on a main depend above eighty,000, displays exceptional overall performance for huge versions of merging stars,” De Marco stated. “With Octo-Tiger, we cannot only cut down the wait time considerably, but our versions can solution quite a few additional of the issues we care to inquire.”

Octo-Tiger is presently optimized to simulate the merger of effectively-solved stars that can be approximated by barotropic constructions, this kind of as white dwarfs or primary sequence stars. The gravity solver conserves angular momentum to machine precision, thanks to a correction algorithm. This code takes advantage of HPX parallelization, allowing the overlap of get the job done and conversation and main to exceptional scaling attributes to resolve huge problems in shorter time frames.

“This paper demonstrates how an asynchronous undertaking-centered runtime technique can be employed as a realistic substitute to Message Passing Interface to assist an vital astrophysical dilemma,” Diehl stated.

The analysis outlines the current and planned places of advancement aimed at tackling a number of actual physical phenomena related to observations of transients.

“While our individual analysis fascination is in stellar mergers and their aftermath, there are a wide variety of problems in computational astrophysics that Octo-Tiger can deal with with its essential infrastructure for self-gravitating fluids,” Motl stated.

The animation (https://www.youtube.com/watch?v=hg9MQNLLJw4) was geared up by Shiber, who claims: “Octo-Tiger demonstrates exceptional overall performance both of those in the accuracy of the remedies and in scaling to tens of hundreds of cores. These results demonstrate Octo-Tiger as an perfect code for modeling mass transfer in binary techniques and in simulating stellar mergers.”